U.S. patent number 8,803,264 [Application Number 13/118,275] was granted by the patent office on 2014-08-12 for room-temperature magnetoelectric multiferroic thin films and applications thereof.
This patent grant is currently assigned to University of Puerto Rico. The grantee listed for this patent is Ram S Katiyar, Ashok Kumar, James F Scott. Invention is credited to Ram S Katiyar, Ashok Kumar, James F Scott.
United States Patent |
8,803,264 |
Katiyar , et al. |
August 12, 2014 |
Room-temperature magnetoelectric multiferroic thin films and
applications thereof
Abstract
The invention provides a novel class of room-temperature,
single-phase, magnetoelectric multiferroic
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x (0.2.ltoreq.x.ltoreq.0.8)
(PFW.sub.x-PZT.sub.1-x) thin films that exhibit high dielectric
constants, high polarization, weak saturation magnetization, broad
dielectric temperature peak, high-frequency dispersion, low
dielectric loss and low leakage current. These properties render
them to be suitable candidates for room-temperature multiferroic
devices. Methods of preparation are also provided.
Inventors: |
Katiyar; Ram S (Guaynabo,
PR), Kumar; Ashok (San Juan, PR), Scott; James F
(Cambridge, GB) |
Applicant: |
Name |
City |
State |
Country |
Type |
Katiyar; Ram S
Kumar; Ashok
Scott; James F |
Guaynabo
San Juan
Cambridge |
PR
PR
N/A |
US
US
GB |
|
|
Assignee: |
University of Puerto Rico (San
Juan, PR)
|
Family
ID: |
51267273 |
Appl.
No.: |
13/118,275 |
Filed: |
May 27, 2011 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
61348895 |
May 27, 2010 |
|
|
|
|
Current U.S.
Class: |
257/421; 438/3;
257/E21.002 |
Current CPC
Class: |
H01L
41/16 (20130101); H01L 29/516 (20130101); H01L
41/318 (20130101); H01L 41/316 (20130101) |
Current International
Class: |
H01L
29/82 (20060101) |
Field of
Search: |
;257/421,E21.002,E29.323
;438/3 |
Primary Examiner: Payen; Marvin
Attorney, Agent or Firm: Hoglund & Pamias, P.S.C. Rios;
Roberto J.
Government Interests
GOVERNMENT INTEREST
The claimed invention was made with U.S. Government support under
grant numbers W911NF-05-1-0340, W911NF-06-1-0030, and
W911NF-06-1-0183 awarded by the Department of Defense (DOD) and
grant number FG 02-08ER46526 awarded by the Department of Energy
(DoE). The government has certain rights in this invention.
Claims
We claim:
1. A (PFW.sub.x-PZT.sub.1-x) magnetoelectric multiferroic composite
comprising: (PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x where,
0.2.ltoreq.x.ltoreq.0.8.
2. The composite of claim 1, wherein 0.2.ltoreq.x.ltoreq.0.6.
3. The composite of claim 1, wherein x is 0.2.
4. The composite of claim 1, wherein said composite comprises a
single-phase.
5. The composite of claim 1, wherein said composite has magnetic
switching of 20-60 .mu.C/cm.sup.2 electric polarizations at a
temperature of about 296 K.
6. The composite of claim 3, wherein said composite has a positive
temperature coefficient of resistivity.
7. The composite of claim 1, wherein said composite has a remnant
polarization in the range of about 20-60 .mu.C/cm.sup.2.
8. The composite of claim 1, wherein said composite has a
dielectric loss in the range of about 1-5%.
9. The composite of claim 1, wherein said composite has a high
dielectric constant in the range of about 1200-3000.
10. The composite of claim 1, wherein said composite has a weak
saturation magnetization in the range of about 0.48-4.53
emu/cm.sup.3.
11. A chemical solution method of fabricating a thin film of
composite of claim 1, said method comprising: providing a
platinized silicon substrate; spin coating a
(PFW.sub.x-PZT.sub.1-x) solution on said substrate; pyrolyzing the
spin coated substrate at a temperature in the range of about
673-773 K; and repeating the steps of spin coating and pyrolyzing
to obtain a film having thickness in the range of about 300-500
nanometers.
12. The method of claim 11, wherein said platinized silicon
substrate is a Pt/TiO.sub.2/SiO.sub.2/Si(100) layered
substrate.
13. The method of claim 11, wherein said (PFW.sub.x-PZT.sub.1-x)
solution has a concentration in the range of about 0.2-0.3 moles
per liter.
14. The method of claim 11, further comprising the step of rapid
thermal annealing at temperatures in the range of about 873-1023 K
for about 60-300 seconds.
15. The method of claim 14, wherein the step of annealing is
carried out in at least two steps of annealing at a temperature of
about 973 K for about 30 seconds followed by annealing at about 923
K for about 150 seconds.
16. A pulsed laser method of fabricating a thin film of composite
of claim 1, said method comprising: providing ceramic targets of
(PFW.sub.x-PZT.sub.1-x)(0.20.ltoreq.x.ltoreq.0.80); depositing a
(PFW.sub.x-PZT.sub.1-x) layer from the ceramic target on a
substrate; annealing the deposited (PFW.sub.x-PZT.sub.1-x) layer at
about 873 K for about 30 minutes in oxygen at a pressure of about
300 Torr; and cooling the annealed layer to about 296 K at a rate
of about 5 K per min.
17. The method of claim 16, wherein the (PFW.sub.x-PZT.sub.1-x)
layer is deposited from the ceramic target at a temperature of
about 873 K under an oxygen pressure of about 200 mTorr, using a
laser energy density in the range of about 1.0-1.5 J/cm.sup.2 at a
repetition rate of about 5-10 Hz.
18. The method of claim 16, wherein the substrate is selected form
the group consisting of: MgO (100),
(LaAlO.sub.3).sub.0.3--(Sr.sub.2AlTaO.sub.6).sub.0.7 (100) or
SrTiO.sub.3.
19. The method of claim 16, wherein the substrate is covered with a
conducting layer of La.sub.0.67Sr.sub.0.33CoO.sub.3 (LSCO) having a
thickness in the range of about 60-90 nm.
20. The method of claim 16, wherein the total thickness of the film
of composite is at least 350 nm.
21. The composite of claim 1, wherein said composite is used in at
least one of: magnetic field sensors, and capacitors.
22. A memory device structure comprising at least one layer of the
composite of claim 1, wherein said layer is sandwiched between at
least one layer of a high-k dielectric material and a metal
electrode.
23. The memory device of claim 22, comprising a memory window
ranging from about 2-4 V and a leakage current density in the range
of about 10-9 to bout 10-6 A/cm.sup.2.
24. The memory device of claim 22, wherein said high-k dielectric
material comprise at least one of: SrTiO.sub.3, HfO.sub.2,
DyScO.sub.3, and LaGdO.sub.3.
25. The memory device of claim 22, wherein said metal electrode
comprises one of: Pt and Au.
26. The memory device of claim 22, wherein the layer of high-k
dielectric is about 2-8 nm thick.
27. A method of fabricating a memory device structure comprising:
providing a p-type Si substrate; depositing at least one layer of a
high-k dielectric material on said substrate, wherein the
dielectric layer is about 2-8 nm thick; and depositing at least one
layer of a (PFW.sub.x-PZT.sub.1-x) composite of composition,
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x where,
0.2.ltoreq.x.ltoreq.0.8, on said dielectric layer, wherein the
composite layer is about 350 nm thick.
28. The method of claim 27, wherein the high-k dielectric material
is selected from a group comprising at least one of: SrTiO.sub.3,
HfO.sub.2, DyScO.sub.3, and LaGdO.sub.3.
29. The method of claim 27, wherein the high-k dielectric layer is
about 2-8 nm thick.
30. The method of claim 27, wherein said dielectric layer is
deposited on said substrate by pulsed laser deposition at a
temperature in the range of about 973-1073 K in an oxygen
atmosphere at a pressure in the range of about 50-100 mT.
31. The method of claim 27, wherein the step of depositing
(PFW.sub.x-PZT.sub.1-x) composite layer on the dielectric layer
comprises: irradiating the composite layer with a laser radiation
under an oxygen pressure of about 200 mTorr at a temperature of
about 873 K; annealing the irradiated composite layer at a
temperature of about 873 K for about 30 min in an oxygen atmosphere
at a pressure of about 300 Torr; and cooling to room temperature at
a rate of about 5 K/min.
Description
BACKGROUND OF THE INVENTION
The development of practical magnetoelectric (ME) multiferroic
devices for industry is concentrated on two embodiments: The first
is for weak-field sensors for applications such as mine detection
in harbors or related military devices. In this case practical
devices favor sandwich-type bilayer or multilayer composites of
ferroelectrics such as lead zirconate-titanate (PZT) with magnets
such as terphenyl-d but this configuration yields low ME voltage
coefficients which is not suitable for weak field sensor
applications. The second is for random access memory elements where
the ability to switch magnetization with applied electric fields
(and vice versa) would permit the combination of fast (possibly
sub-nanosecond) WRITE operation with nondestructive magnetic READ
operation. Such a combination would produce formidable competition
for FLASH EEPROMs, particularly in view of the fact that
magnetoelectric RAMs would operate at <1.0V, an international
target for all microelectronics in the next decade which FLASH
devices are unlikely to meet without cumbersome internal charge
pumps.
However, most multiferroics to date (such as terbium manganites
TbMnO.sub.3 or TbMn.sub.2O.sub.5) switch only nC/cm.sup.2 with
applied magnetic fields--approximately 1000.times. too small for
reliable discrimination between "1" and "0" states. Moreover, they
operate only at cryogenic temperatures. BiFeO.sub.3 might operate
at room temperature, but it exhibits a weak magnetism (ca. 0.03
emu/cc), high leakage current, and it does not permit switching of
neither magnetization with electric fields nor polarizations with
magnetic fields. There are several other magnetically frustrated
systems that have been identified which show gigantic
magnetoelectric effects at low temperature. Among them, MnWO.sub.4
shows external magnetic field dependence of its ferroelectric loops
below room temperature. Similarly, high-temperature multiferroic
behavior has been reported in PbFe.sub.0.5Ti.sub.0.5O.sub.3 and in
PbFe.sub.2/3W.sub.1/3O.sub.3/PbTiO.sub.3. Others have studied the
relaxor properties of PFW and PFW-PT ceramics.
Clearly a new line of thinking is required to overcome the
intrinsic limitations of direct linear magnetoelectric (ME)
coupling of form PM in the free energy, where P is polarization and
M, magnetization. There have been recent efforts by researchers
throughout the world to produce electric switching of
magnetizations or magnetic switching of polarizations in
multiferroic magnetoelectrics. Generally, schemes for switching
from +M to -M with electric field (E) or conversely from +P.sub.r
to -P.sub.r with magnetic field (H) have been examined. In the
present state of the art there is a need for a new single-phase
material that can exhibit magnetoelectricity (not necessarily
linear) at room temperature.
SUMMARY OF THE INVENTION
The present invention provides a novel class of magnetoelectric
multiferroic compositions that have a room-temperature magnetic
switching of large electric polarizations with high memory window
in MMIS structure paving way for non-volatile memory
applications.
According to an aspect of the invention, a novel class of
room-temperature, single-phase, magnetoelectric multiferroic
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.4703).sub.1-x (0.2.ltoreq.x.ltoreq.0.8)
(PFW.sub.x-PZT.sub.1-x) thin films are provided that exhibit high
dielectric constants (1200-3000), high polarization (22-60
.mu.C/cm.sup.2), weak saturation magnetization (0.48-4.53
emu/cm.sup.3), broad dielectric temperature peak, high-frequency
dispersion, low dielectric loss and low leakage current.
According to a further aspect of the invention, the inventive films
are suitable for room-temperature multiferroic devices. Near 20% of
PFW, the ferroelectric relaxor state is nearly unstable at room
temperature against long-range ferroelectricity which in turn gives
a large magnetoelectric effect: switching not from +Pr to -Pr with
applied H, but from Pr to zero with applied H of less than a
Tesla.
According to another aspect of the invention, a Metal Multiferroic
Insulator Semiconductor (MMIS) device structure is provided on
Pt/(PFW.sub.x-PZT.sub.1-x)(0.2.ltoreq.x.ltoreq.0.4)/SrTiO.sub.3(6
nm)/Si (111) substrate showing a well behaved and saturated C(V)
hysteresis with 3-4 V memory window, and low leakage current
density (10.sup.-8-10.sup.-6 A/cm.sup.2) @ 150 kV/cm.
According to one aspect of the invention, the synthesis and
fabrication of a novel class of room temperature, single-phase,
magnetoelectric multiferroic thin films of composition
(PFW.sub.x-PZT.sub.1-x)(0.2.ltoreq.x.ltoreq.0.8) is provided
showing magnetic control of large polarization (.about.22-60
.mu.C/cm.sup.2) merely less than a Tesla.
According to another aspect of the invention, the method of
fabricating magnetoelectric multiferroics of composition
(PFW.sub.x-PZT.sub.1-x) (0.2.ltoreq.x.ltoreq.0.8) that exhibit high
dielectric constants, weak saturation magnetization, broad
dielectric temperature peak, high-frequency dispersion, low
dielectric loss and low leakage current is provided.
According to still another aspect of the invention,
(PFW.sub.x-PZT.sub.1-x) composite thin films are provided, where
x=0.20; exhibiting a giant polarization of 60 .mu.C/cm.sup.2,
moderate magnetization 6-8 emu/cc, and a high piezoelectric
coefficient at room temperature.
According to one aspect of the invention, a method of fabricating
(PFW.sub.x-PZT.sub.1-x) (x=0.20) composite thin films that switch
large polarizations from 22 .mu.C/cm.sup.2 to 0 .mu.C/cm.sup.2
under the application of 0.5 Tesla is provided.
According to another aspect of the invention, the device
incorporating the above films permit 3-state logic (+P.sub.r, O,
-P.sub.r) and a condenser with >5000% magnetic field change in
its capacitance; in H=0 the coercive voltage is 1.4 V across 300 nm
for +P.sub.r to -P.sub.r switching, and the coercive magnetic field
is 0.5 T for +P.sub.r to zero switching.
According to a further aspect of the invention, a method of
fabricating (PFW.sub.x-PZT.sub.1-x) (x=0.20) composite thin films
in which increasing magnetic field lengthens the relaxation time by
.times.500 from <200 ns to >100 .mu.s is provided, wherein it
strongly couples the polarization relaxation and spin
relaxations.
According to one aspect of the invention, it provides a family of
room-temperature, magnetoelectric mutiferroic devices fabricated
and designed by tailoring compositions of (PFW.sub.x-PZT.sub.1-x)
(0.2.ltoreq.x.ltoreq.0.8) at the "B-site" that show high dielectric
constant, low dielectric loss, diffused ferroelectric phase
transition, low leakage current and high temperature dependent
dielectric tunability.
According to another aspect of the invention, a memory device
structure incorporating the (PFW.sub.x-PZT.sub.1-x)
(x=0.20.ltoreq.x.ltoreq.0.80) composite and a method of fabrication
of such device is provided.
According to still another aspect of the invention, a
(PFW.sub.x-PZT.sub.1-x) (x=0.20.ltoreq.x.ltoreq.0.80) composite
that is used in microelectronic applications in the form of
magnetic field sensors, and capacitors is provided.
According to a further aspect of the invention, a MMIS device
structure of Pt(PFW.sub.x-PZT.sub.1-x)/SrTiO.sub.3/Si (111) is
provided, fabricated for different compositions of
(PFW.sub.x-PZT.sub.1-x) (x=0.20.ltoreq.x.ltoreq.0.60), showing well
behaved and saturated C(V) hysteresis with 3-4 V memory windows,
and low leakage and serving as excellent candidates for
non-volatile memory applications and multiferroic RAMs.
BRIEF DESCRIPTION OF THE DRAWINGS
Further features and advantages of the invention will become
apparent from the following detailed description taken in
conjunction with the accompanying figures showing illustrative
embodiments of the invention, in which:
FIG. 1 shows a plot of Room temperature magnetization versus
magnetic field (M-H) response of the (PFW.sub.x-PZT.sub.1-x)
(x=0.20 (a), x=0.30 (b), and x=0.40 (c)) thin films, according to
the present invention.
FIG. 2 shows a plot of Three-state logic switching (+P, O,
-P.sub.r) in 0.2 PFW/0.8PZT, according to the present
invention.
FIG. 3 shows a plot of Polarization vs. electric field hysteresis
loops for sol-gel derived (PFW.sub.x-PZT.sub.1-x) (x=0.20 (a),
x=0.30 (b), and x=0.40 (c)) at 300 K, 200 K and 150 K respectively,
according to the present invention.
FIG. 4 shows a plot of Fatigue behavior of the films with 100 kHz
bipolar square waves in the presence of 350 kV/cm external electric
field over 109 cycles, according to the present invention.
FIG. 5 shows a plot of Real and imaginary part of dielectric
permittivity as function of frequencies under the application of
external magnetic field from 0 to 0.85 T, according to the present
invention.
FIG. 6 shows a Cole-Cole plot of dielectric data under the action
of applied magnetic field, according to the present invention.
FIG. 7 shows a plot of Relaxation time as a function of applied
magnetic field, according to the present invention.
FIG. 8 shows a plot of Vogel-Fulcher-type fitting of relaxation
frequency, according to the present invention.
FIG. 9 shows a plot of Room temperature polarization vs. electric
field hysteresis loops for PLD grown (PFW.sub.x-PZT.sub.1-x)
(x=0.20 to 0.40 with very high polarization, well saturation and
low coercive field, according to the present invention.
FIG. 10 shows plots of Temperature dependent PLD grown
(PFW.sub.x-PZT.sub.1-x) (x=0.20), according to the present
invention.
FIG. 11 shows plots of Room temperature magnetization versus
magnetic field (M-H) response of the PLD grown
(PFW.sub.x-PZT.sub.1-x) (x=0.20) thin films (a) in plane
magnetization, (b) out of plane i.e. magnetization along c-axis,
according to the present invention.
FIG. 12 shows a schematic diagram of a fabricated MMIS device
structure, according to the present invention.
FIG. 13 shows a plot of High frequency C(V) characteristics of
Pt/(PFW.sub.x-PZT.sub.1-x)/SrTiO.sub.3/(111)Si MMIF structures for
different compositions of PFWx: (a) x=0.20, (b) x=0.30, (c) x=0.40,
according to the present invention.
FIG. 14 shows plots of Room temperature J(E) characteristics of
Pt/(PFW.sub.x-PZT.sub.1-x)/SrTiO.sub.3/(111)Si MMIF structures as
function of sweep gate field >150 kV/cm (>15MV/m) for
different compositions of PFW, according to the present
invention.
Throughout the figures, the same reference numbers and characters,
unless otherwise stated, are used to denote like elements,
components, portions or features of the illustrated embodiments.
The subject invention will be described in detail in conjunction
with the accompanying figures, in view of the illustrative
embodiments.
DETAILED DESCRIPTION OF THE INVENTION
The present invention relates to a novel room-temperature,
single-phase, magnetoelectric multiferroic structure of composition
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x where,
0.2.ltoreq.x.ltoreq.0.8. All the terms used have the same meaning
as is used in general aspect of the art, unless otherwise
specified. As used hereinafter, the following terms shall have the
following definitions: "Magnetoelectric" refers to the effect of
induced dielectric polarization of a material in an applied
magnetic field or an induced magnetization in an external electric
field; "Piezoelectricity" refers to the ability of materials to
generate an electric field in response to an applied mechanical
force; "Relaxor" refers to materials having naturally occurring
short-range polar nano-ordered regions and highly
frequency-dependent dielectric properties; "Remanent polarization,
P.sub.r" refers to the spontaneous polarization under the
application of external field of zero bias; "Coercive E-field,"
E.sub.c refers to the electric field which is used to switch the
polarization to generate the different logic state for memory
applications. The E-field frequency is 50 Hertz for the purpose of
this description; "E.sub.max," refers to the maximum electric field
that is used for the hysteresis loop measurements; "Leakage current
density" (J'') refers to the current per unit area, measured as
function of applied electric field; "Polarization fatigue P.sub.sw"
refers to the loss in polarization after certain number of cycles
of device utility; "B-site": The general formula for perovskite
oxides is ABO.sub.3, where specific elements occupy the A and B
sites in the crystal lattice, and O is oxygen. For
(PFW.sub.x-PZT.sub.1-x), Pb is on the A-site and Zr, Ti, Fe, and W
share the B-site; "Room temperature" refers to a temperature of 296
K.; "Positive temperature coefficient of resistivity" refers to the
increase in resistance with increase in temperature; "Negative
differential resistivity" refers to the decrease in resistance with
increase in applied external voltage.
The composite (PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x is referred to as
PFW.sub.x-PZT.sub.1-x throughout this description. The present
invention provides a class of single-phase, magnetoelectric
multiferroic materials comprising at least one
piezoelectric/ferroelectric composition and one
magnetic/ferroelectric relaxor composition, wherein the
piezoelectric/ferroelectric composition is PZT and the
magnetic/ferroelectric relaxor composition is PFW.
The invention provides a composite
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.4703).sub.1-x where, 0.2.ltoreq.x.ltoreq.0.8
that exhibits ferroelectricity, piezoelectricity and (weak)
ferromagnetism above room temperature with magneto-electric
coupling via strain mediated electrostriction and magnetostriction.
The composite exhibits a wide range of unique properties including
but not limited to room temperature magnetic switching of large
electric polarizations; remanent polarization in the range of 22-60
.mu.C/cm.sup.2; dielectric loss in the range of 1-5%; high
dielectric constant in the range of 1200-3000; and a weak
saturation magnetization. The invention also provides a
(PFW.sub.x-PZT.sub.1-x) (x=0.20.ltoreq.x.ltoreq.0.80) composite
that exhibits a weak saturation magnetization in the range of
0.48-4.53 emu/cm.sup.3 that shows an increase of 15 emu/cm.sup.3
with increase in Fe and W compositions. A preferred embodiment of
the present invention provides a composite
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.0.2
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.0.8 that exhibits a giant
polarization of 60 .mu.C/cm.sup.2, moderate magnetization 6-8
emu/cc, a high piezoelectric coefficient at room temperature, and a
positive temperature coefficient of resistivity and a negative
differential resistivity.
Another embodiment of the invention provides a
(PFW.sub.x-PZT.sub.1-x) (x=0.20.ltoreq.x.ltoreq.0.80) composite
that is used in microelectronic applications as magnetic field
sensors, and capacitors. The composite exhibits remarkable change
in capacitance >5000% under the application of small magnetic
field (0.5 T) rendering it suitable for use as magnetic field
sensor and capacitor.
The invention provides a chemical solution method of fabricating a
thin film of composite (PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x where,
0.2.ltoreq.x.ltoreq.0.8. The method is carried out by providing a
platinized silicon substrate; spin coating a
(PFW.sub.x-PZT.sub.1-x) solution on said substrate; pyrolyzing the
spin coated substrate at a temperature in the range of 673-773 K;
and repeating the steps of spin coating and pyrolyzing to obtain a
film having thickness in the range of 300-500 nanometers. In a
preferred embodiment of the invention, the platinized silicon
substrate is a Pt/TiO.sub.2/SiO.sub.2/Si(100) layered substrate. In
this chemical solution method, high purity (>99.9%) precursor
materials, lead acetate trihydrate, iron 2-4 pentanedieonate,
titanium isopropoxide, zirconium isopropoxide and tungsten
isopropoxide are used to prepare the (PFW.sub.x-PZT.sub.1-x)
solution of concentration in the range of 0.2-0.3 moles per liter.
The invention provides a chemical solution method of fabricating a
thin film of composite (PFW.sub.x-PZT.sub.1-x)
(0.2.ltoreq.x.ltoreq.0.8) that further includes the step of rapid
thermal annealing at temperatures in the range of 873-1023 K for a
time period in the range of 60-300 seconds. Rapid thermal annealing
is first carried out at a temperature of 973 K for 30 seconds
followed by annealing at 923 K for 150 seconds.
The invention can be alternatively carried out by a pulsed laser
method of fabricating a thin film of composite
(PFW.sub.x-PZT.sub.1-x) (0.2.ltoreq.x.ltoreq.0.8). The method
comprises providing ceramic targets of (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.80); depositing a (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.80) layer from the ceramic target on a
substrate; annealing the deposited (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.80) layer at 873 K for 30 minutes in oxygen
at a pressure of 300 Torr; and cooling the annealed layer to 296 K
at a rate of 5 K per min.
The invention also provides a pulsed laser method of fabricating a
thin film of composite
(PFW.sub.x-PZT.sub.1-x)(0.20.ltoreq.x.ltoreq.0.80), wherein the
composite layer is deposited from the ceramic target to the
substrate at a temperature of 873 K under an oxygen pressure of 200
mTorr, using a laser energy density in the range of 1.0-1.5
J/cm.sup.2 at a repetition rate of 5-10 Hz. The substrate is MgO
(100), (LaAlO.sub.3).sub.0.3--(Sr.sub.2AlTaO.sub.6).sub.0.7 (100)
or SrTiO.sub.3. The substrate is covered with a conducting layer of
La.sub.0.67Sr.sub.0.33CoO.sub.3 (LSCO) having a thickness in the
range of 60-90 nm. The total thickness of the film is at least 350
nm.
All the composites having composition
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x wherein,
0.2.ltoreq.x.ltoreq.0.8 prepared by the pulsed laser method are
weakly ferromagnetic and ferroelectric at some temperature, but the
increasing amounts of PZT raises the relaxor-to-ferroelectric phase
transition from about 150K at 60% to 300K at 80%, so the 80% PZT
specimens combine the most interesting properties at ambient
temperatures. They exhibit well saturated hysteresis loops with
high remanent polarization (P.sub.r) of about 55-35 .mu.C/cm.sup.2
at an external electric field of 550 kV/cm at room temperature; the
remanent polarization decrease with increasing Fe and W
compositions. Further enhancement in the Fe and W compositions
magnetic Neel's temperature and Curie temperature (ferroelectric
phase transition temperature) come near to room temperature which
is best for magnetoelectric device fabrication.
The magnetic hysteresis loops of composites with 20% PFW, 30% PFW,
and 40% PFW is shown in FIG. 1. The M-H data display weak
ferromagnetism. The induced magnetization for
(PFW.sub.x-PZT.sub.1-x) was 5 emu/cc at a high magnetic field of
1.5 T; up to this external magnetic field the M-H curve for 20% PFW
shows an unsaturated loop with a low saturation magnetization
(M.sub.s.about.0.48 emu/cc) and a low coercive field (H.about.0.12
T), but at 40% the saturation is better and M.sub.s=4.5 emu/cc. A
super-exchange in the disordered regions through
Fe.sup.+3--O--Fe.sup.+3 is expected to yield antiferromagnetic
ordering but at the low Fe concentrations in this material, weak
ferromagnetism is probably caused by spin clustering.
A preferred embodiment of the present invention provides a
single-phase tetragonal composite
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x (0.2.ltoreq.x.ltoreq.0.4)
prepared by the chemical solution and pulsed laser methods onto
Pt/TiO.sub.2/SiO.sub.2/Si(100) and SrTiO.sub.3/Si substrates.
The present invention provides a memory device structure that
includes at least one layer of the composite
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x wherein,
0.2.ltoreq.x.ltoreq.0.8, preferably 0.2.ltoreq.x.ltoreq.0.6, more
preferably 0.2.ltoreq.x.ltoreq.0.4; that is sandwiched or layered
between at least one layer of a high-k dielectric material and a
metal electrode. The invention can be embodied in a non-volatile
memory device structure that includes at least one layer of the
composite (PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x wherein,
0.2.ltoreq.x.ltoreq.0.8, preferably 0.2.ltoreq.x.ltoreq.0.6, more
preferably 0.2.ltoreq.x.ltoreq.0.4; that is sandwiched or layered
between at least one layer of a high-k dielectric material and a
metal electrode. The layer of high-k dielectric material is
selected from a group comprising SrTiO.sub.3, HfO.sub.2,
DyScO.sub.3, and LaGdO.sub.3 and the metal electrode is Pt or Au.
The thickness of the high-k dielectric layer is in the range of 2-8
nm. The memory device structure displays a memory window ranging
from 2-4 V and a low leakage current density in the range of
10.sup.-9 to 10.sup.-6 A/cm.sup.2.
The invention provides a method of fabricating a memory device
structure that includes the steps of providing a p-type Si
substrate; depositing at least one layer of high-k dielectric
material on the substrate, wherein the dielectric layer is 2-8 nm
thick; and depositing at least one layer of composite
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x where,
0.2.ltoreq.x.ltoreq.0.8, on the said dielectric layer, wherein the
composite layer is 350 nm thick. The high-k dielectric material is
selected from a group including but not limited to SrTiO.sub.3,
HfO.sub.2, DyScO.sub.3, and LaGdO.sub.3. The layer of dielectric
material has a thickness in the range of 2-8 nm and it is deposited
on the p-type Si substrate by pulsed laser deposition at a
temperature in the range of 973-1073 K in an oxygen atmosphere at a
pressure in the range of 50-100 mT. In a preferred embodiment of
the invention, the step of depositing a layer of the composite
(PbFe.sub.0.67W.sub.0.33O.sub.3).sub.x
(PbZr.sub.0.53Ti.sub.0.47O.sub.3).sub.1-x where,
0.2.ltoreq.x.ltoreq.0.8 on the dielectric layer, comprises:
irradiating the composite layer with a laser radiation under an
oxygen pressure of 200 mTorr at a temperature of 600.degree. C.;
annealing the irradiated composite layer at a temperature of 873 K
for 30 min under an oxygen pressure of 300 Torr; and cooling to
room temperature at a rate of 5 K/min.
The invention also provides room-temperature, single-phase,
magnetoelectric multiferroic (PFW.sub.x-PZT.sub.1-x)
(0.2.ltoreq.x.ltoreq.0.8) thin films for non-volatile memory
applications in FERAMS. A preferred embodiment of the invention
provides a room-temperature, single-phase, magnetoelectric
multiferroic (PFW.sub.x-PZT.sub.1-x) (0.2.ltoreq.x.ltoreq.0.6) thin
films for non-volatile memory applications in FERAMS.
FIG. 12 shows a memory device according to an embodiment of the
invention. The memory device comprises a p-type Si substrate
deposited with at least one layer of
(PFW.sub.x-PZT.sub.1-x)(0.2.ltoreq.x.ltoreq.0.6) composite; at
least one layer of high k-dielectric material, wherein the high-k
dielectric material is selected from a group including but not
limited to SrTiO.sub.3, HfO.sub.2, DyScO.sub.3, and
LaGdO.sub.3.
The composite (PFW.sub.x-PZT.sub.1-x) (0.20.ltoreq.x.ltoreq.0.80)
of the present invention have high potential for non-volatile
memory and multiferroic RAM applications.
EXAMPLE 1
Preparation and Properties of (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.40) by Chemical Solution Method
Materials and Methods
Ferroelectric (PFW.sub.x-PZT.sub.1-x) (0.20.ltoreq.x.ltoreq.0.40)
thin films were deposited on Pt/TiO.sub.2/SiO.sub.2/Si(100)
substrates using chemical solution deposition (CSD). The high
purity (>99.9%) precursor materials (i.e., lead acetate
trihydrate, iron 2-4 pentanedieonate, titanium isopropoxide,
zirconium isopropoxide and tungsten isopropoxide) were bought from
Alfa Aesar and used for making stock solutions. For good quality
films a (PFW.sub.x-PZT.sub.1-x) (0.20.ltoreq.x.ltoreq.0.40)
solution of 0.2-0.3 molar concentration was spin coated at 3000 rpm
for 30 seconds and pyrolyzed at 673 K for 2 minutes. This process
was repeated several times in order to get desired film thickness.
Rapid thermal annealing (RTA) was performed at different
temperatures from 873 K-1023 K for different time intervals
.about.60-300 seconds to get the desired phase and high density.
The substrate consists of a layer of Si (about 350 .mu.m thick),
SiO.sub.2 (about 0.6 .mu.m thick), TiO.sub.2 (about 0.1 .mu.m
thick) and Pt (about 0.15 .mu.m thick). The films are deposited on
a substrate which is approx. 300 nm thick.
DC sputtering was carried out for depositing the Pt top electrode
of 3.1.times.10.sup.-4 cm.sup.2 area using a shadow mask. The
dielectric properties in the frequency range of 100 Hz to 1 MHz
were studied using an impedance analyzer HP4294A over a wide range
of temperature attached to a temperature controlled probe station.
Magnetic properties were investigated using a vibrating sample
magnetometer (VSM). The polarization versus electric field (P-E)
hysteresis loop of the capacitor was measured using the RT 6000HVS
ferroelectric tester (Radiant technology) operating in virtual
ground mode.
Results
X-ray diffraction patterns (XRD) of the (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.40) thin films deposited on
Pt/TiO.sub.2/SiO.sub.2/Si substrates at 673 K and post-annealed at
1023 K for 60 seconds, continuously further annealed at 923 K for
180 seconds for better homogeneity and crystallinity. In order to
get the desired phase, several temperatures were chosen for RTA;
the compositions listed provide the best single-phase compound at
each annealing temperature. The XRD analysis indicates that the
films grown at 973 K were single-phase polycrystalline in nature
with less than 1% impurities (pyrochlore) phases. The synthesized
complex single-phase perovskite possesses more than five elements
in the perovskite octahedra, which makes it very unlikely to get a
100% pure phase, but the electrical properties of all the films
having less than 5% pyrochlore were excellent, giving very similar
polarizations and electrical responses.
These exhibit ferroelectricity and (weak) ferromagnetism above room
temperature with strain coupling via electrostriction and
magnetostriction. Application of modest magnetic field strength
(.mu..sub.0H<1.0 Tesla) destabilizes the long-range
ferroelectric ordering and switches the polarization from ca. 22
.mu.C/cm.sup.2 (0.22 C/m.sup.2) to zero (relaxor state) as shown in
FIG. 2. This offers the possibility of three-state logic (+P, 0,
-P) and magnetically switched polarizations. Because the switching
is of large magnitude (unlike the very small nC/cm.sup.2 values in
terbium manganites) and at room-temperature, commercial devices
should be possible.
The electric-field-induced polarization switching (P-E) behavior
was studied at low frequency at 50 Hz using Sawyer-Tower
measurements. FIG. 3 shows the ferroelectric hysteresis loops for
(PFW.sub.x-PZT.sub.1-x) (x=0.20 (a), x=0.30 (b), and x=0.40 (c)) at
300 K, 200 K and 150 K respectively. The films exhibit well
saturated hysteresis loops with remanent polarization (P.sub.r) of
about 22, 20, and 11 .mu.C/cm.sup.2 and coercive field of 47 kHz,
64 kHz, and 111 kHz respectively for (a), (b) and (c) under 400
kV/cm applied external electric field. Enhancement in remanent
polarization may be attributed to the relaxed local strain (due to
larger grains) and/or greater compositional. In order to check the
real-time polarization behavior of (PFW.sub.x-PZT.sub.1-x) (x=0.20)
films and thus their utility in memory devices, fatigue tests were
performed. Fatigue behavior of (PFW.sub.x-PZT.sub.1-x) thin films
was studied with 100 kHz bipolar square waves in the presence of
350 kV/cm external electric field. The first fatigue data points
were obtained after 10.sup.5 cycles and continued until 10.sup.9
cycles, as shown in FIG. 4. There was a modest polarization loss
(<12%). Twelve percent decay in fatigue of
(PFW.sub.x-PZT.sub.1-x) (x=0.20) thin films on a platinized silicon
substrate after 10.sup.9 cycles much better than the early reports
for PZT, indicating its suitability for memory device
applications.
Three-state logic and switching at room temperature is shown in
FIG. 2. +P.sub.r to -P.sub.r switching is observed at E.sub.c=48
kV/cm (1.4 V across 300 nm); and magnetic switching from +Pr to
zero is shown at 0.5 Tesla. In microelectronics memory
applications, it is of greater interest to switch magnetization M
with electric field E than to switch polarization P with magnetic
field H. The device can be placed in a dc field of H=0.49 T with
only 100 Gauss required to switch. Three-state logic at a switching
field of 100 Gauss could be very useful.
The polarization response at 0.5 T is shown in the insert in FIG.
2; representing a leaky linear dielectric. The reason for this
behavior is shown clearly in the plot of dielectric constant versus
frequency at different magnetic fields shown in FIG. 5. At H=0.80
T, the dielectric constant is 1430 for f=1 kHz but it is 233 at
f=100 kHz. This is simply due to the field dependence of the
relaxation time, and the fact that as H increases the size
(correlation length) of spin clusters in this magnetic relaxor
increases and hence their fluctuation time decreases (in an
extrapolated infinite field H, the system would have long-range
ferromagnetic order and the relaxation frequency would go to zero).
The fact that this strongly influences their dielectric
susceptibility requires strong coupling between spin fluctuations
and polarization fluctuations. FIG. 5 shows that the peak of the
dielectric loss at H=0.85 T is at f=ca. 2-3 kHz. Note that this
agrees with the inflection frequency of 2-3 kHz estimated from real
part of permittivity, suggesting a magnetic field-induced Debye
relaxation. The relaxation time obtained from the characteristic
peak in dielectric loss spectra from FIG. 5 and/or from the
stretched exponential parameters of FIG. 6 are plotted in FIG. 7.
It indicates that increasing H lengthens the relaxation time by
500.times. from <200 ns to >100 .mu.s, and it couples
strongly the polarization relaxation and spin relaxations.
These characteristic frequencies can be accurately fitted to high
precision the Vogel-Fulcher-type equation:
.times..function. ##EQU00001##
where H.sub.f is a freezing field and replaces the analogous term
in T.sub.f (freezing temperature) in the usual Vogel-Fulcher
Equation; .mu..sub.B is the Bohr magneton; and H.sub.R, the
relaxation field. Eq. 1a arises from an E.sup.2H.sup.2 term in the
free energy and is independent of the sign of H. The fitted data
provide freezing field H.sub.c=0.92.+-.0.07 Tesla and
characteristic frequency f.sub.0=40.+-.1 MHz as shown in FIG.
8.
Conclusions
A bigger family of room-temperature single-phase magnetoelectric
multiferroic (PZT-PFW).sub.x thin films has been synthesized by CSD
techniques which shows a high dielectric constant, low dielectric
loss, highly frequency-dispersive susceptibility, more than 50%
temperature dependent dielectric tunability, and
above-room-temperature dielectric maxima. They exhibit a
ferroelectric polarization of 22 .mu.C/cm.sup.2 and low coercive
field of 48 kV/cm with 12% fatigue in polarization after 10.sup.9
cycles, thus suggesting a potential candidate for memory
applications. At room temperature a weak
ferromagnetic/antiferromagnetic M-H hysteresis was detected which
suggests an extra degree of freedom of Fe ions in complex
octahedra, tilting of Fe.sup.+3--O--Fe.sup.+3 linkage, and charge
ordering with nearest neighbors. The ferroelectric hysteresis flops
under the application of external magnetic field (>0.50 T),
revealing a strong magnetoelectric coupling in the present system,
but this is not direct, rather it is indirect through strain via
electrostriction plus magnetostriction. It is not due to a
conventional magnetically-induced phase transition (instead, due to
a field-dependent spin relaxation time). This new magnetoelectric
effect, switching from +P.sub.r to zero with applied magnetic
field, is 1000 times greater than in rare-earth manganites and
occurs at room temperature, suggesting a variety of new
magnetoelectric devices, including very large magneto-capacitance
and three-state logic elements.
EXAMPLE 2
Preparation and Properties of (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.80) by Pulsed Laser Method
Materials and Methods
Stoichiometric ceramic targets of
(PFW.sub.x-PZT.sub.1-x)(0.20.ltoreq.x.ltoreq.0.80) with 10% excess
of lead oxide (to compensate for loss of volatile Pb) were
synthesized by a conventional solid-state route. Three different
compositions of (PFW.sub.x-PZT.sub.1-x) films with varying
deposition temperature were fabricated by pulsed laser deposition
(PLD) employing a KrF excimer laser (A=240 nm). The conducting
La.sub.0.67Sr.sub.0.33CoO.sub.3 (LSCO) layer was grown on MgO (100)
substrate at 873 K under an oxygen pressure of 300 mTorr, using a
laser energy density of (1.8 J/cm.sup.2) and repetition rate of 10
Hz, followed by normal cooling in oxygen atmosphere. LSCO was used
as bottom electrode for all kind of electrical measurements. The
(PFW.sub.x-PZT.sub.1-x) layer was then deposited on the LSCO layer
at 873 K under an oxygen pressure of 200 mTorr, using a laser
energy density of (1.5 J/cm.sup.2) and repetition rate of 10 Hz.
After deposition, (PFW.sub.x-PZT.sub.1-x) was annealed at 873 K for
30 minutes in oxygen at a pressure of 300 Torr. Finally, the films
were cooled down to room temperature at a slow rate. The total
thicknesses of the films were around 350 nm. The structural
analysis was done with a Siemens D500 x-ray diffractometer (Cu
K.alpha. radiation) in a .theta.-2.theta. scan. Atomic force
microscopy (Veeco-AFM-contact mode) was used to examine the
morphology and the surface roughness. The film thickness was
determined using an X-P-200 profilometer. The electrical and room
temp magnetic measurements were carried out with the same
equipments as for CSD films. Temperature dependent magnetic
measurements were carried out using a Quantum Design MPMS XL-7
SQUID magnetometer.
Results
XRD patterns of (PFW.sub.x-PZT.sub.1-x) films indicate epitaxial
(cube by cube) growth along (100) directions. The peak positions
were shifted towards higher Bragg angles with increase in Fe and W
compositions. The full width at half maximum (FWHM) increases
slightly for higher PFW percentages, indicating the local strain
contribution to the FWHM is reduced by the presence of a unique
crystalline orientation. Cell parameters of these films were
estimated due to the limited number of peaks in XRD spectra showed
tetragonal crystal structure till x.about.0.50. Lattice mismatch
between the substrate and the films were analyzed on the basis of
the in-plane epitaxial relationship (PFW.sub.x-PZT.sub.1-x)
[100].parallel.LSCO [100].parallel.MgO [100]; the lattice mismatch
was calculated using the equation
##EQU00002## where .di-elect cons. is the lattice mismatch,
.alpha..sub.sub is the lattice constant the substrate and,
.alpha..sub.film is the lattice constant at the composite. A
tensile strain (4.5-4.9% along the a-axis; 2.3-2.9% along the
c-axis) and compressive strain (-4.9 to -4.5% along the a-axis,
-7.3 to -6.6% along the c-axis) was observed between the bottom
electrode and the dielectric composite layer. Two
oppositely-directed strains across the substrate-film and
film-bottom electrode interfaces suggest qualitatively that
in-plane compressive gradient strain (inhomogeneous strain) exists
in these films, which may cause higher ferroelectric phase
transition temperatures in PLD grown thin films, compared with the
CSD specimens.
The electric-field-induced polarization study was carried out at
very low operating frequency .about.50 Hz using Sawyer-Tower
circuit. FIG. 9 shows the ferroelectric hysteresis loops of PLD
grown (PFW.sub.x-PZT.sub.1-x) (x=0.20 to 0.40) at room temperature
(297 K) respectively. These films exhibit well saturated hysteresis
loops with high remanent polarization (P.sub.r) of about 55-35
.mu.C/cm.sup.2 at 550 kV/cm external electric field, the remanant
polarization decrease with increasing Fe and W composition.
The temperature and frequency dependence of (PFW.sub.x-PZT.sub.1-x)
(x=0.20) are shown in FIG. 10 which shows frequency-independent
dielectric peaks (.di-elect cons..sub.max) at 550 K from 100 Hz to
10 kHz and a 40 K shift in .di-elect cons..sub.max temperature for
higher frequencies (up until 1 MHz). These observations reveal
strain-induced relaxor ferroelectric properties. The tangent loss
shows the character typical of relaxor materials: loss increases
with increase in applied frequency. The frequency dispersion in the
dielectric constant does not follow that of classical relaxor
materials, but instead is more similar to that of impurity-induced
relaxor materials like Ba.sub.1-xSr.sub.xTiO.sub.3 and
SrBi.sub.2Nb.sub.2O.sub.9. SrBi.sub.2Nb.sub.2O.sub.9 can be tuned
as a relaxor ferroelectric by doping with Ba, Sr, or Ce ions. A
huge frequency dispersion in the dielectric spectra below and above
the ferroelectric phase transition is noticed with extra energy
storage capacity during the heating and cooling process. Similar
phenomena were observed in the loss-tangent spectra. Heating and
cooling effects on the dielectric constant and loss spectra at
selected frequencies (10 kHz and 100 KHz) are shown in FIG. 10, (b)
and (d).
The most striking features of increased Fe and W compositions in
(PFW.sub.x-PZT.sub.1-x) are as follows: (i) the ferroelectric phase
transition temperature shifts towards room temperate; (ii)
ferroelectric relaxor behavior (e.g., frequency dispersion)
increases; (iii) energy storage capacity increases; (iv) more
temperature hysteresis for T.sub.c upon heating and cooling; (v)
loss tangent increases with increase in ratio of Fe and W
compositions in solid solutions of (PFW.sub.x-PZT.sub.1-x).
The magnetization versus applied field (M-H) characterization of
(PFW.sub.x-PZT.sub.1-x) (x=0.20) films performed for in-plane and
out-of-plane field directions is shown in FIG. 11. The induced
magnetization values were between 6-8 emu/cc at 4 kOe magnetic
field. M(H) along the c-axis showed well saturated ferromagnetic
hysteresis with 2-3 emu/cc remanent magnetization and 230 Oe
coercive field. The in-plane M(H) curve displayed a small opening
(hysteresis) H.sub.0=2500e coercive field. The origin of magnetism
can be explained in light of the presence of d.sup.n ions at B-site
which may provide in-plane ferromagnetism. The magnetization of
(PFW.sub.x-PZT.sub.1-x) (x=0.20) is better and well defined compare
to same composition prepared by CSD techniques; this may be due to
inhomogeneous (gradient) compressive strain in the films which
cause a small distortion in the Fe.sup.+3--O--Co.sup.+3 spin angle,
producing room-temperature ferromagnetism via spin canting.
Super-exchange in the disordered regions through
Fe.sup.+3--O--Fe.sup.+3 is expected to yield antiferromagnetic
ordering; there is a possibility of strong magneto-electric
coupling, suggesting the converse effect of our earlier
report--viz., electric fields E controlling the magnetic properties
at both mesoscopic and microscopic scale.
Conclusions:
A large family of room-temperature magnetoelectric mutiferroic
devices has been designed at by tailoring compositions of
(PFW.sub.x-PZT.sub.1-x) at the B-site. It is found that with
suitable compositions, both ferroelectric and ferromagnetic phase
transition temperatures can be made to coincide near room
temperature, which will be of great interest for devices and basic
physics. Ferroelectricity diminished whereas relaxor behavior and
anti/ferromagnetism improved with increase in FeW/ZrTi ratios. The
temperature of the dielectric maximum temperature shifted strongly
to lower temperatures with increase in Fe and W percentages. The
ferroelectric phase transition temperatures in PLD films are higher
than that of CSD specimens with the same nominal compositions, most
likely attributable to in-plane gradient compressive strain. The
ferro/antiferromagnetic/magnetic relaxor properties of PLD films
are better that those CSD films. The basic properties of these
multiferroics promise a bright future for room-temperature
magnetoelectrics having both ferroic phase transitions at ambient
temperature, despite the fact that their magnetoelectric coupling
is nonlinear.
EXAMPLE 3
Fabrication of MMIS Device Structure of (PFW.sub.x-PZT.sub.1-x)
(0.20.ltoreq.x.ltoreq.0.40) on p-Type Silicon
Materials and Methods:
To fabricate the MMIS structure, 6 nm SrTiO.sub.3 (ST) was
deposited on a p-type Si(111) substrate by PLD technique. Attempts
at fabrication of the structure by depositing amorphous SrTiO.sub.3
at low temperatures yielded much higher leakage currents, in turn
giving unsaturated accumulation regions under the application of
negative bias voltage. So instead crystalline SrTiO.sub.3 (6 nm)
was grown at 973 K with 100 mTorr of oxygen atmosphere. Three
different compositions of (PFW.sub.x-PZT.sub.1-x) films with
varying deposition temperature were fabricated by PLD employing a
KrF excimer laser (.lamda.=240 nm). The (PFW.sub.x-PZT.sub.1-x)
layer was then deposited on the ST layer at 600.degree. C. under an
oxygen pressure of 200 mTorr, using a laser energy density of (1.5
J/cm.sup.2) and repetition rate of 10 Hz. After deposition the
(PFW.sub.x-PZT.sub.1-x) was annealed at 873 K for 30 minutes in
oxygen at a pressure of 300 Torr. Finally the films were cooled
down to room temperature at a rate of 5 K/min. The total
thicknesses of the films were around 350 nm. All other structural
and electrical characterizations were done similar to previous
example.
Results:
Crystal structure of (PFW.sub.x-PZT.sub.1-x) thin films on high
k-SrTiO.sub.3 (6-nm) coated p-Si (111) substrate showed a good
single-phase tetragonal material of perovskite structure. The MMIS
structure fabricated in the above example can be represented by
FIG. 12 and for the present example the high k dielectric shown in
the figure is SrTiO.sub.3. FIGS. 13 (a), (b), (c) show memory
window data as functions of sweep voltage at 1 MHz operating
frequency. Curve (a) is for 20% PFW; (b) is 30% PFW; (c) is 40%
PFW. A large window of ca. 4.0V is observed for the 40% PFW sample.
The capacitance (C)-voltage (V) curve is similar to that of a
normal metal-oxide-semiconductor (MOS) structure, clearly showing
regions of accumulation, depletion, and inversion. It should be
remember that all these compositions show very large polarization
and low coercive field, as presented in our earlier reports. The
figure displays large clockwise C-V hysteresis and a memory window
controlled by ferroelectric hysteresis of the
(PFW.sub.x-PZT.sub.1-x) on the Si surface. The memory window of
(PFW.sub.x-PZT.sub.1-x) as function of sweep voltage with an almost
negligible memory window below -3 V sweep voltage, saturated C-V
hysteresis until -10V, above which gate voltage the MMIS structure
started leaking during the DOWN sweep, whereas during the UP sweep
(in gate voltage) accumulation regions showed good saturation.
These observations indicate that the materials have high potential
for RAM applications.
FIG. 14 shows the current density vs applied electrical field J(E)
data. The specimens, as in earlier figures, are: (a) 20% PFW; (b)
30% PFW; (c) 40% PFW. All these specimens showed very low leakage
current for this MMIS structure; these leakage currents are higher
when the electric field is applied from the semiconductor p-Si
side. Different specimens showed the same characteristic J(E)
behavior of very low leakage .about.10.sup.-5 to 10.sup.-7
A/cm.sup.2 (10.sup.-9 to 10.sup.-11 A/m.sup.2) under -10V gate
voltage. Due to low leakage behavior, it is hard to distinguish
leakage current differences among them; but at x=0.30 of PFW has
less leakage. J(E) analysis suggest that the conduction mechanism
either due to Poole-Frenkel (PF) or Schottky emission (SE).
Conclusions:
MMIS device structures of Pt/(PFW.sub.x-PZT.sub.1-x)/SrTiO.sub.3/Si
(111) are fabricated for different compositions of PFW: (a) x=0.20,
(b) x=0.30, (c) x=0.40. All these composition showed well behaved
and saturated C(V) hysteresis with 3-4 V memory windows, suggesting
good candidates for multiferroic RAMs. J(E) characteristics of
these specimens illustrate low leakage currents at room temperature
dominated by bulk-limited n-type Poole-Frenkel conduction. Energy
band match-up diagram and experimental results suggest that all
these specimens might provide a wide memory window and good DOWN
saturation if used with alternative high band-gap and high
k-dielectric materials instead of SrTiO.sub.3. Since all the
compounds are multiferroic in nature, under the application of
external magnetic field it may be possible to destabilize the
saturated gate voltage and produce more than two logic states with
additional functionalities.
Although the invention has been described in conjunction with
specific embodiments, it is evident that many alternatives and
variations will be apparent to those skilled in the art in light of
the description. Accordingly, the invention is intended to embrace
all of the alternatives and variations that fall within the spirit
and scope of the appended claims.
* * * * *